On Mon, 17 Apr 2000, Manfred Spraul wrote:
>Could you give more details?
I'm debugging Mario's random lockup with `sar` in background since some
time now, I found last week the offender (serial proc read) and I fixed
the crash a few days ago:
ftp://ftp.*.kernel.org/pub/linux/kernel/people/andrea/patches/v2.2/2.2.15pre17/serial-proc-1
Posted the above patch to Mario mailbox a few days ago to give it a try
and he reported his machine runs rock solid now.
I simply rewrote the whole serial read-proc logic from scratch in a sane
way. I had not yet the time to look on what gone wrong in the old very
_insane_ implementation (that's the only reason I didn't yet posted the
thing to l-k since I wanted to find out exactly the bug of the old code
first!). I could reproduce weird thing with the old implementation. If the
output in your system is not long enough try with bs=10 or so.
The crash was subtle to find because a cat reads with buffers 4096 wide,
and only a dd bs=xx could exploit it.
So far I only had a very superficial look to the old implementation and I
noticed probably this was wrong in the old implementation:
*start = page + (begin-off);
I need a good sleep now :), more thoughts on the above line next morning.
Andrea
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sun Apr 23 2000 - 21:00:12 EST